better compressor pedal??

[Hides-His-Eyes]

Well, I do have an oscilloscope, so if you want to tell us more about it...
How should it be biased?

Fair enough! You need to disable the side chain so the FET can be manually controlled.

Easiest way is to short the 100k feedback resistor on U1b.
Set postgain to zero resistance.
Set ratio to zero resistance.
Input a suitable test tone; level is not important provided it doesn't clip. Say, 100mVpk, 1kHz, ish.
Monitor the input and output waves and adjust the bias until the output amplitude is equal to the input.

That's it! For anyone without a scope they'd have to switch back and forth between the effect and bypass, and adjust the bias by ear for equal loudness.

I will try and make a sound clip this weekend.

Using a boutique mindset...

Could you just make it a pot and call it a feature? ("volume"?)

[mr_deadmaxxx]

i think i have decided to go with an Ibanez CP9..is this a good compressor?something that can be all-around (all-genre) ?

[merlinb]

To anyone it may interest, I have uploaded a PCB layout for my compressor:
http://s81.photobucket.com/albums/j207/merlinblencowe/Compressor/

The layout is for the "light" version, which has only level and threshold controls (see associated schem) and electronic switching.

To bias the FET, set the footswitch to bypass mode. Input a suitable test signal and adjust bias until output signal is equal to input signal.

[askwho69]

Wow thanks a lot! can i ask is this a quite compressor ? always asking this hehehe!

[auden100]

Thanks for the layout Merlin. I've been working on a perf layout for the 'full' version for several days now (harder than I thought), and I noticed you changed a couple of things for the light version. Most of it, obviously, is for the purpose of streamlining the controls, but the others I was wondering if you preferred in the 'light' configuration. Namely, the C2 100p cap on U1d, and  the reversing of R17 and C7. The extra power filtering makes sense. I don't know much about electronics, but is it safe to assume that the added BC337s are part of a buffered bypass control?

I figure the answer to all of these questions is 'yes', but I'd rather ask than assume.

Thanks for sharing your circuit, and I appreciate the help. I'm looking forward to eventually getting this thing together.

[askwho69]

ill give this a shot next @ but im a bit confused on the layout? no signs i dont know ctrl usage? is it the switch? for momentary?

[merlinb]

Namely, the C2 100p cap on U1d,

That just limits the upper bandwidth of the amplifier stage; standard practice. Doesn't affect it audibly.

and  the reversing of R17 and C7.

That was to make layout easier. Since they're in series, it doesn't matter which way round those components go.

I don't know much about electronics, but is it safe to assume that the added BC337s are part of a buffered bypass control?

Yep. Q2 disables the side chain and Q3 shorts the level pot, returning the gain to unity during bypass mode.

[auden100]

Got around to finishing this. Unverified, but I think it should work. May still try to clean things up a bit. This was one tricky circuit to layout. If anyone sees an error, let me know. I combined the revisions in the Light version with the controls in the Full version. There wasn't an object that looked like the trimmers that I use (from futurlec), so I substituted with the three vero cut symbols.

Roughly, 7cm by 4cm. Should fit comfortably in a 1590BB.

[Nich]

I am a fan of my Pigtronix Philosopher's Tone.  It adds what you need and lets you blend your dry signal with the effect nicely so you don't lose your original sound.  Plus it has a rad analog dist which doesn't work for everyting but has its place for sure. 
The MXR Sustain Unit looks pretty interesting too. 

[askwho69]

Got around to finishing this. Unverified, but I think it should work. May still try to clean things up a bit. This was one tricky circuit to layout. If anyone sees an error, let me know. I combined the revisions in the Light version with the controls in the Full version. There wasn't an object that looked like the trimmers that I use (from futurlec), so I substituted with the three vero cut symbols.

Roughly, 7cm by 4cm. Should fit comfortably in a 1590BB.

now this is a clear layout@

[auden100]

Thanks. Used the new DIYLC V3 Beta. Exported to PDF, and then converted to 300dpi jpeg. Getting very clear results that way.

[thedefog]

i think i have decided to go with an Ibanez CP9..is this a good compressor?something that can be all-around (all-genre) ?

I've found in my experience with compressors is that there is never really one all-around one that will work great with everything and every style. The CP9 is a great design, as it blends the top end of your original signal with the compressed signal. It'll probably work very well for most styles and genres.

[askwho69]

Auden100 can you make it in pcb ? i wanna try your  layout soon

[BubbaFet]

Sheesh... Merlin is not only the 'Valve Wizard',
he is an 'Effect Wizard' as well ! I'll bet he has
one of the seven 'Electronic Rings of Power'.

We're all lucky that he is a GOOD wizard.

So Merlin, when does your Effects Book get released?
Soon, I hope, as I've got $40 burning a hole in my pocket.
(It will be devalued to $30 by this summer... please hurry!)

[Ripdivot]

I have had the "light" version of Merlin's compressor on my breadboard for about a week now. Over all it is a pretty nice compressor but I had to make some changes to get it to work with my gear and the components I have on hand. First off I didn't have any J112 Fets so I tried what I have which is J201, 2N5457 and 2N5458. I also had to change R9 (4K7) to 470 ohms and the threshhold pot from 50K to 5K in order to get a decent range of control. I tried several of each of the fets I have and found that the J201 made the threshold way too touchy. The 2N5457 was a little better and the 2N5458 was the best of the three. I couldn"t get much compression at all with the threshold at 50K and R9 at 4K7. Merlin (if you are out there) I assume these changes I had to make are all do to the VGSoff of the fet being incorrect?

[merlinb]

Merlin (if you are out there) I assume these changes I had to make are all do to the VGSoff of the fet being incorrect?

Yes almost certainly. The J201 needs less than -1.5V to turn off- very sensitive!
However, I am surprised that you needed to reduce the threshold resistor, because that would increase the gain of side chain, yet that FET needs less voltage to drive than my version! Did you use a scope to bias the FET, or ears?

EDIT: Ack! Just noticed that D1 and D2 are both backwards in my schematics! I am updating them now. This would explain your problem, sorry! 

[merlinb]

Here's a vero layout (unverified). Doesn't include the electronic switching, it's for DIYers with old skool bypass.
BOM here: http://s81.photobucket.com/albums/j207/merlinblencowe/Compressor/

[Ripdivot]

Merlin, so I changed the diodes around so they are placed correctly and reinstalled the original components and it works OK. I still had to drop the threshold resistor down to 1K to get a better compression range. This must still be because of the incorrect fet. As it stands now (with the 2N5458 Fet) my circuit works best with a 25K threshold pot and a 1K threshold resistor. I will try and source a J112 this week and report back with my findings. It works nice in the subtle settings but gets pretty hissy with more compression. Sometimes I wish I could get a lot more squash out of it...

[merlinb]

It works nice in the subtle settings but gets pretty hissy with more compression.Sometimes I wish I could get a lot more squash out of it...

That seems to suggest a problem. With the side chain gain at max, such that you need to tunr the level way up to hiss levels, you should be getting mammoth amounts of squish!
Here are some initial characteristics I measured. x-axis is input signal in mV(p-p) and y-axis is output in mV(p-p) with level at minimum.
All curves are with  with max ratio, except where noted.
With everything dimed the knee starts around 20mVp-p input, levelling off at about 20dB compression, which is serious squish!

[Ripdivot]

Here are the numbers I am getting (25K threshold pot and 1K threshold resistor, 2N5458 Fet):

with controls at minimum

-30db in / -31db out
-20db in / -23dbout
-10db in / -15db out
0db in / -6db out

with threshold maxed and gain at minimum

-30db in / -37db out
-20db in / -30db out
-10db in / -24db out
0db in / -14db out